Story
Nov 6, 2026
OpenAI Offers Grid Relief, but Voters Still See a Data Center Threat
OpenAI is promising power cutbacks, public planning and no NDAs with officials as opposition to AI data centers spreads. The company says flexible campuses can help the grid; critics remain focused on bills, pollution and lost local control.
Opposition had been building well before OpenAI’s latest charm offensive. Across the summer, communities on both sides of the political divide challenged enormous AI data-center proposals, worried that the facilities would raise utility bills, worsen pollution and arrive with scant public input. In Utah, residents protested a project expected to consume more than twice the state’s projected 2025 peak electricity demand; in Texas, researchers warned planned projects could erase decades of air-quality gains within five years.1
On Tuesday, that unease showed up at the ballot box. Voters supported restrictions on, or rejected pro-development measures for, 40 of 41 local data-center ballot questions across 10 states, according to an analysis cited by CNBC.2 The result was a blunt political warning to an industry racing to secure the computing capacity behind new AI models.
Two days later, OpenAI answered with a policy blueprint for data centers it owns and operates. It said it would seek community input, publish construction plans, avoid non-disclosure agreements with public officials and cover the full cost of powering its sites.2 Chris Lehane, the company’s chief global affairs officer, called the proposal an effort to establish “a very high floor in terms of how people move forward.”2
OpenAI had also laid out a more operational pledge: its Grid Protectors Program, piloted at a planned campus in Effingham County, Georgia. The company says it could cut up to one gigawatt of demand during extreme weather or grid emergencies—roughly the electricity use of 700,000 homes.3 John McCarrick, OpenAI’s head of energy policy, said the industry was moving away from the assumption that data centers require near-perfect, uninterrupted availability.3
That is the company’s central pitch: steady AI demand can spread grid costs, while emergency flexibility can ease peak strain. But opponents are asking a harder question—whether promises of relief will outweigh the concrete burdens of giant campuses. Several states have already acted to stop companies from shifting energy costs onto residents, while moratoriums and other restrictions continue to gather force.1